We have developed a new method to discover microbial causes of human disease, sequence-based computational subtraction. In this method, sequences from diseased tissue are compared to the human genome computationally, and the filtered sequences are highly enriched for non-human nucleic acids. I propose to apply computational subtraction to search for viruses that cause lymphomas associated with immunodeficiency, most notably post-transplant lymphoproliferative disorder and HIV-associated lymphoma. First, I propose to use specimens of post-transplant lymphoproliferative disorder, known to be positive for Epstein-Barr virus, to refine our methods for library generation and sequencing. In particular, we would like to test the use of normalization, subtraction, and concatenation techniques. Once we have improved these techniques, I plan to focus on searching for novel viruses in immunodeficiency-associated lymphomas of unknown etiology. We plan to generate cDNA libraries from immunodeficiency-associated lymphoma biopsy specimens, to sequence a sampling of these libraries, and then to subtract the sequences computationally and experimentally against the human genome. Filtered sequences will be tested further for specific association with lymphoma using the polymerase chain reaction. Should we successfully identify novel lymphoma-associated sequences, we will then attempt to generate molecular clones of the entire putative viruses and begin to characterize the protein products of their genomes. Computational subtraction is a broadly applicable method. While we will begin our pathogen discovery projects in cancer, our methods will be broadly applicable to many human diseases. These include auto-immune diseases and inflammatory diseases, as well as uncharacterized epidemics, whether natural or bio-terrorist in origin.